“I feel like it’s giving me a lot as a language teacher to be a learner myself”: Factors affecting the implementation of a multilingual pedagogy as reported by teachers of diverse languages

To boost the foreign language learning process, language teachers need to know how to implement a multilingual pedagogy, that is, they should be able to draw on their and their students’ knowledge of other languages during lessons. This qualitative study explored the extent to which 21 foreign language teachers in Norwegian and Russian upper-secondary schools were willing and able to implement multilingual teaching practices and the factors that they thought affected this implementation. The findings revealed three main factors, namely, their language knowledge, their positioning as language learners, and the level of support they received, which the participants reported as strongly influencing the extent to which they were able and willing to draw on their and their students’ multilingualism as a pedagogical resource. The findings also indicated that participants did not implement multilingual teaching practices differently based on the languages they taught, although there were differences between the participants from Norway and Russia concerning the teaching of English. The study has important implications for research on language teaching and learning in multilingual environments, educational institutions, and teacher development programs.


Introduction
Governments around the world are implementing initiatives in schools and universities that encourage younger generations to learn multiple foreign languages (FLs) so that individuals can develop a deeper understanding and appreciation of the cultural and linguistic diversity that surrounds them and acquire the skills to secure themselves professionally and personally in a globalized world (Calafato, 2021a;2021b;Gao & Zheng, 2019). As part of these initiatives, the ministry of education in some countries, such as in Norway (UDIR, 2013), for example, encourage FL teachers, most of whom know multiple languages (i.e., they are multilingual) (Calafato, 2019), to implement multilingual teaching practices (MTPs) during lessons. MTPs are an umbrella term for practices where teachers systematically draw on their and their students' whole language repertoire as a pedagogical resource (Gopalakrishnan, 2020). MTPs can include translanguaging, where students and teachers use their languages integratively and interconnectedly (see García et al., 2017), activities that combine receptive skills (e.g., reading) in one language with productive skills (e.g., speaking) in another (Calafato, 2019;2020a), and the awakening to languages approach, where teachers seek to raise their students' general awareness of linguistic diversity (Candelier, 2004). The use of MTPs has been shown to boost student performance across all skill levels when compared to the use of a target language-only approach (Brown, 2021) and finds support among both students and parents (de Figueiredo, 2011;Moody et al., 2019). MTPs also emphasize the dynamic, interconnected, and compounding nature of language learning (Cook, 2016;Jessner, 2008), signaling a shift away from the fractional view of multilingualism that sees languages as disconnected entities in the mind (i.e., multiple monolinguals in one person) (Grosjean, 1989). In light of the multilingual initiatives promoted by governments and the emphasis on using MTPs in FL teaching, it becomes important to understand the extent to which FL teachers can implement these initiatives, what they think about the use of MTPs, and the extent to which they are able and willing to implement them.
Thus far few studies have explored FL teachers' reported or observed implementation of MTPs (Calafato, 2019;Lorenz et al., 2021), and fewer still (Aslan, 2015;Calafato, 2021b) have explicitly investigated the factors that may affect their use of such practices. A review of studies by Calafato (2019) on multilingual FL teachers and their identity, including their beliefs and practices, found that many studies had explored what teachers specifically thought about their students' multilingualism and whether they would be willing to use their students' home languages during lessons. Fewer studies had investigated what FL teachers thought about their own multilingualism, how they related to the languages they taught, the extent to which they reported (or were observed) using MTPs during lessons, and what activities they implemented when doing so. In this latter group of studies, it was discovered that the implementation of MTPs by FL teachers was unsystematic and they often avoided using MTPs due to beliefs in the Chomskyan ideal of the native speaker (Lee, 2016), lack of teaching experience and maladaptive training (Vaish, 2012;Zheng, 2017), and inability to reflect on and verbalize language form, function, and meaning (i.e., metalinguistic knowledge) (Aslan, 2015). Specific examples of MTPs used by the participants in the reviewed studies included translation, translanguaging, drawing cross-cultural contrasts, and language comparison (e.g., concerning cognates, grammar, utterances, etc.). The findings from the review have been borne out in subsequent studies, with FL teachers being found to implement MTPs in an ad hoc manner and, at times, appearing to subscribe to a fractional view of multilingualism (Gopalakrishnan, 2020;Lorenz et al., 2021).
At the same time, what has remained somewhat constant in most studies is that the focus continues to be, as already mentioned, on what FL teachers think about specifically their students' multilingualism rather than how they see themselves as multilingual language users and teachers, and how their own needs, experiences, and competences affect their teaching. Indeed, while an exploration of what FL teachers think about their students' multilingualism can provide some insights into their beliefs about language learning and use, it alone may ultimately reveal little about factors that influence their implementation of MTPs. This is because such practices might be more strongly linked to how they see themselves rather than how they see their students. This is implied in several studies where the FL teacher participants were found to implement MTPs only after reflecting on their own identity as multilingual language users (Ng, 2018;Park, 2012) and not that of their students. It is also worth emphasizing here that the multilingual initiatives launched by governments around the world, whether in Asia (Calafato, 2021a;Gao & Zheng, 2019), Europe (Calafato, 2020a;European Council, 2019), or the Middle East (Tang & Calafato, 2021), encourage students to learn multiple FLs as part of their school or university education, often two or more simultaneously. Yet, studies on the implementation of MTPs by FL teachers have seldom explored this multiplicity of languages, focusing instead purely on one language, mostly English (Calafato, 2019), with less attention paid to languages other than English (LOTEs) (Aslan, 2015;Calafato, 2020a;2021b;Brown, 2021;Haukås, 2019;Tang & Calafato, 2021). The result is that we have an incomplete picture of the use of MTPs in FL teaching, one that does not accurately reflect the compound state of FL education in schools and universities, where students are encouraged to not only draw on their whole language repertoire but also to do so while learning multiple FLs.
A few studies have also revealed variation in teachers' implementation of MTPs depending on the languages they teach, with those teaching English implementing MTPs to a statistically significantly lesser extent than those teaching LOTEs (Calafato, 2021b). However, we know little for certain regarding why such disparities exist given the very limited number of studies that have compared teaching practices across languages (e.g., Aslan, 2015;Calafato, 2021b;Jiang et al., 2014). Meanwhile, the greater emphasis placed by governments on the learning of multiple FLs in schools and universities has led to instances where teachers teach two or more FLs. The teaching of two or more FLs may affect FL teachers in different, more complex ways than if they only taught one FL given how dynamically languages interact in the mind (Jessner, 2008), with far-reaching implications for teachers' cognition and teaching practices. For example, studies on monolingual, bilingual, and trilingual students indicate that they process and learn new languages differently in terms of speed and strategies (Festman, 2021;Jordà, 2017). One could hypothesize, therefore, that teachers who teach more than one FL may not only implement MTPs more frequently but may also teach each language in their repertoire using different strategies. Much like in their multilingual students, there could manifest considerable variation in how they relate to the languages they teach. Since only a handful of studies have reported on how the teaching of multiple FLs can affect teachers (Aslan, 2015;Wernicke, 2018), there is little empirical data to draw on to provide conclusive evidence in support of these claims.
This article reports on a study that sought to address the above-mentioned gaps in our knowledge concerning FL teacher multilingualism. The study investigated the factors that FL teachers in Norwegian and Russian upper-secondary schools reported as influencing their implementation of MTPs and how they, as multilingual language users, related to the languages in their teaching repertoire. The participants were teachers of Chinese, English, French, German, and Spanish, and many taught two FLs concurrently. In investigating their implementation of MTPs, the study also explored the participants' language backgrounds, their rapport with their students and colleagues, their thoughts concerning the school FL curriculum in their respective countries, and their approach to teaching languages.

Shining the spotlight back on language teachers
To ensure the success of multilingual initiatives, such as the ones formulated by the education ministries in Norway and Russia (Minobr, 2018;UDIR, 2013), where students can learn two FLs in schools and the FL curriculum stresses the importance of being multilingual to navigate a globalized world, policymakers should first obtain a holistic understanding of the FL teachers that are tasked with implementing them. This means understanding how they see themselves in relation to the languages they teach, as well as how their multilingualism and other micro-, meso-, and macro-level factors (e.g., government and school support for multilingualism) affect their views about multilingualism and approach to teaching. Signaling a reorientation of language learning and teaching research, the transdisciplinary and identity frameworks proposed by the Douglas Fir Group (2016) and Gee (2001) encourage researchers to consider the multifaceted nature of language learning and teaching at the micro-, meso-, and macro-levels, where belief systems, cognitive capacities, and identities, among myriad other elements, intermingle. The Douglas Fir Group (2016), for instance, stresses that the "competence of multilingual speakers is the holistic sum of their multiple-language capacities" (p. 26) and that the "languages and cultural schemata of a multilingual interact" (p. 30) in complex ways as new languages are learned. This interaction of various language and cultural schemata has been explored by researchers mostly in students and not in teachers, yet it is just as likely that multilingualism influences teachers in equally, if not more complex ways than the students they teach.
In adopting a more holistic approach to exploring teacher multilingualism, especially as this concerns their ability and willingness to implement MTPs, it is also important to understand how teachers position themselves vis-à-vis the languages they teach and their multilingualism. Positioning can be described as "situating oneself or others with particular rights and obligations through conversation," and has largely been studied in relation to language learners rather than teachers (Kayı-Aydar, 2019, p. 5). According to Deppermann (2015, p. 372), positioning provides a window into an individual's "practical, emotional, and epistemic commitment to identity-categories and associated discursive practices." Regarding the implementation of MTPs, for example, FL teachers might position themselves as lacking competence or experience in implementing these in one or more languages. They may, as already mentioned, subscribe to the nativespeaker ideal that views the monolingual native speaker as the ideal speaker hearer of a given language and behave accordingly when teaching (Lee, 2016). If they teach multiple FLs, they may position themselves differently per language (Aslan, 2015), something which has rarely been explored. By investigating how teachers position themselves and the extent to which various micro-, meso-, and macro-level factors impact their ability and willingness to implement MTPs, we can obtain deeper insights into how they make pedagogical decisions and how they relate to the languages they teach. Such insights are of vital importance to policymakers, teacher educators, and educational institutions (as well as for teachers and students) in light of the initiatives taken by governments around the world to promote multilingualism among the populace through FL education in schools and universities, where greater emphasis is placed on the need to validate and harness the multilingualism of both students and teachers through MTPs.

Multilingualism, multicompetence, and affordances
In this study, multilingualism is defined as an individual's use of more than one language (for a discussion of variations in multilingualism, see Cenoz, 2013), without a lower limit being set on their linguistic competence (Lehmann, 2007). To better conceptualize how languages can interact in the mind of FL teachers and how they may draw on these to implement MTPs, one can turn to the theories of multicompetence (Cook, 2016) and affordances (Aronin, 2014). Multicompetence represents the overall state of a multilingual individual's mind (Cook, 2016), including how their languages interact with each other and the extent to which they use their languages integratively or separately. In FL teachers, their level of multicompetence could determine the extent to which they implement MTPs during lessons in that some might use only the target language, thereby opting to teach monolingually, whereas others could use additional languages to varying degrees of activation. As for affordances theory, it posits that each language in a multilingual individual's repertoire provides certain possibilities for action when engaging with their environment so that the more languages one knows, the more possibilities to act one has (Aronin, 2014). For example, a language learner who knows English, Italian and Spanish may learn Portuguese more rapidly than one who knows only English. This is because the former's knowledge of additional languages furnishes them with insights and skills unavailable to the latter, including a larger resource pool of morphosyntactic, social, pragmatic, and lexical competences linked to multiple languages. Multilingualism may provide FL teachers, much like their students, with access to a similar resource pool to draw on when teaching, which may result in their implementation of activities that are more creative, multilingual, and effective.
At present, few studies have explored FL teacher multilingualism and their teaching practices through the lens of multicompetence or affordances theory, which lies in stark contrast to the amount of language learner research carried out using these or similar theories. These latter studies indicate that higher levels of linguistic competence in multiple languages can lead learners to draw more frequently on their multicompetence during interactions (Chang, 2020;Dewaele & Wei, 2012). It is likely, therefore, that FL teachers who have advanced linguistic competence in the languages they teach could draw on their multicompetence more freely during lessons than those who possess weak linguistic competence. Linguistic competence comprises language skills (e.g., reading, writing, etc.), fluency (i.e., how effortlessly one can deploy one's language skills), as well as metalinguistic (i.e., language knowledge) and pragmatic knowledge, among other abilities (Lehmann, 2007), but also covers the many ways in which an individual can use a given language. Among these components, metalinguistic knowledge (i.e., the ability to verbalize language form, function, and meaning) and crosslinguistic awareness, that is, "the awareness (tacit and explicit) of the interaction" between languages (Jessner, 2008, p. 279), including their similarities and differences, are "instrumental for teachers if they are to help learners develop individual multilingualism and to promote plurilingual approaches to teaching languages" (Otwinowska, 2014, p. 101). Otherwise, FL teachers can neither use MTPs systematically nor do they have access to a metalanguage they can teach their students to help them organize and draw on their knowledge of multiple languages. However, only a few studies have explored metalinguistic knowledge and cross-linguistic awareness in FL teachers, with the emphasis being mostly placed on their non-nativeness rather than their multilingualism (for a brief review, see Erlam et al., 2009).

Foreign language education in Norway and Russia
The reasons to focus on Norway and Russia in the study are threefold. First, there is a need for more contrastive analysis within and between countries (e.g., De Angelis, 2011) when it comes to researching FL teachers' implementation of MTPs so that we may obtain a better understanding of how macro-level factors like language policy, language ideologies, and nationality affect this implementation (Douglas Fir Group, 2016;Gee, 2001). Secondly, research on FL teacher multilingualism and the implementation of MTPs is mostly non-existent in the Russian context (Calafato, 2020b(Calafato, , 2021b and remains understudied in the Norwegian one (exceptions to this include Haukås, 2019;Lorenz et al., 2021). Norway and Russia also make for an interesting comparison in that they are neighbors, yet the former is firmly situated in Europe, geographically and culturally, while the latter has a more hybrid identity (Davydova, 2019;Protassova, 2010). Third, the FL curriculum for schools in both Norway and Russia emphasizes the importance of learning multiple FLs and using these languages to boost one's knowledge in other fields. In Norway, the curriculum stresses the need for students to draw on their knowledge of other languages and previous language learning experiences to learn new languages (UDIR, 2013). In Russia, the curriculum highlights the importance of learning languages to navigate a multicultural and multilingual world, one where students must apply their knowledge of languages across subjects and disciplines (Minobr, 2018;PIRAO, 2017). The two countries are also similar in that students have the option to learn two FLs at school (the first one usually being English), choosing their second FL from among French, German, and Spanish (Haukås & Speitz, 2018;Ustinova, 2005). A large number of Russian schools now also offer Chinese, whereas other foreign languages like Arabic, Italian, and Japanese are offered by a very limited number of schools in both countries.
Students in Russia start learning their first FL in Grade 2, followed by a second FL in lower-secondary school (Minobr, 2018). In Norway, students start learning English in Grade 1 and they can learn an additional FL in Grade 8 (Speitz & Lindemann, 2002). By the time students in Norway and Russia enter uppersecondary education, that is, the final years of school before they begin university, their proficiency in multiple FLs is likely at its highest level when compared to any other period from their school years. In theory, this could allow uppersecondary school FL teachers to implement a wider range of MTPs with their students than FL teachers in lower grades due to upper-secondary school students having attained comparatively greater competence in the FLs taught (see Hofer & Jessner, 2019; for a discussion of the link between aspects of linguistic competence and multicompetence and affordances, see Section 2.2.).

Research questions
To shed more light on the implementation of MTPs by teachers of one or more FLs in upper-secondary schools in Norway and Russia, this study explored the following research questions: 1. What factors do the participants consistently report as influencing their implementation of MTPs? 2. Are there differences in the participants' reported implementation of MTPs based on the languages they teach? 3. Are there country-specific differences in the way the participants report implementing MTPs?

Participants
Twenty-one FL teachers participated in the study (10 from Norway and 11 from Russia), which comprised the concluding phase of a larger project on teacher multilingualism and multilingual teaching practices. The previous phases of the project explored the beliefs and practices of multilingual FL teachers in Norway and Russia through quantitative means that provided insights into general trends in both countries (Calafato, 2020a(Calafato, , 2020b(Calafato, , 2021b; however, they lacked the exploratory power and fine-grained analysis afforded by qualitative research methods. As a qualitative component, while this study explored the implementation of MTPs in a small sample of teachers (N = 21), it did so in a way that revealed the specific factors that FL teachers reported as influencing their implementation of MTPs, how they related to the languages they taught, and how they viewed their multilingualism, something that was not explored during the project's quantitative phase. All the participants reported studying at least two FLs in secondary and tertiary education and were non-native speakers of the languages they taught. Ten participants reported teaching only one FL while the rest reported teaching two FLs. The participants were teaching at upper-secondary schools in Norway or Russia at the time of the study. Table 1 provides additional biographical information about those teachers. All names have been anonymized. During this phase of the project, voluntary response sampling was used to recruit participants and drew from the pool of 517 FL teachers that had participated in the earlier phases. These teachers had been recruited through emails sent to upper-secondary schools in the largest cities and counties in Norway (i.e., Oslo, Viken, Vestland, Rogaland, etc.) and Russia (i.e., Moscow, St. Petersburg, Ekaterinburg, etc.) via their listings on official online portals. The emails contained an invitation letter to participate in the study and a questionnaire for potential teacher participants to complete. The teachers were asked to include their contact details at the end of the questionnaire if they wanted to contribute further to the project by participating in interviews. The invitation letter also included details about the project, a description of the participants' rights should they choose to contribute, as well as the contact details of those responsible for the study. Eighty-eight language teachers out of the 517 that completed the questionnaire sent in their email addresses and were subsequently contacted to finalize interview times and dates. Of these, 21 teachers followed through with the interviews, with the others withdrawing from the study due to work constraints brought about by the COVID-19 pandemic.

Data collection and analysis
The study used two rounds of interviews, organized in an open-discussion, mostly unstructured format, to collect data from the participants. The second round of interviews was needed because a preliminary analysis of the data obtained during the first interview round revealed certain themes like the cross-linguistic effects of teaching two FLs that required deeper exploration. In the second round, the interviews were more structured and were held six weeks after the first round.
The open-ended discussion format used in the interviews ensured that the participants could talk freely (Patton, 2014), and it created a more conducive atmosphere for exploring how they positioned themselves (Kayı-Aydar, 2019) regarding the languages they taught, their multilingualism, and their ability and willingness to implement MTPs. Specifically, the participants were asked about their identity as teachers who knew and taught multiple languages, their teaching practices in light of this multiplicity of languages, their language learning experiences, their relationship with their students, colleagues, and the school administration, the FL curriculum, and their experiences teaching in Norway and Russia. The aim in asking these questions was to obtain a more holistic understanding of the extent to which certain micro-, meso-, and macro-level factors (Douglas Fir Group, 2016;Gee, 2001) affected their ability and willingness to implement MTPs while also accounting for their whole language repertoire and relationship with each of the languages therein. In addition, the participants were asked about their years of teaching experience and other biographical information (see Table 1). When conceptualizing the participants' ability and willingness to implement MTPs, including how the languages they knew interacted in their minds and the extent to which they harnessed their multilingualism when teaching, the study drew on multicompetence (Cook, 2016) and affordances theory (Aronin, 2014). It was therefore hypothesized that the more languages the participants knew, the more frequent their reported implementation of MTPs would be and the more diverse their range of activities in this regard.
The interviews were conducted in English, Norwegian, and Russian over Skype, with only one participant being interviewed in person. Each interview lasted around 50 minutes. The interviews were audio-recorded and transcribed (excluding hesitations and pauses), following which each interview transcript was read in detail and broken up into fragments of varying length using an inductive approach that identified recurring themes and ideas through the generation of codes (Braun & Clarke, 2006). The transcripts of the interviews that were conducted in Norwegian and Russian were not translated into English before coding (coding, however, was done in English) and were processed in the original alongside the interviews conducted in English. Data were analyzed using the ATLAS.ti software. Following the generation of an initial set of codes, these codes were used in cross-case analysis (Miles & Huberman, 1994) to identify similarities and differences between the participants concerning how they related to the languages they taught and the factors they reported as affecting their ability and willingness to implement MTPs, including their views about their and their students' multilingualism. This led to a process of refining codes and generating initial themes. The codes and themes generated during this process were subsequently checked against all the interviews until the process reached saturation and a final set of themes and, ultimately, categories, based on micro-, meso-, and macro-level variables, was created.

Findings
The interview data provided notable insights into how the participants felt about their multilingualism, the factors they reported as affecting their ability to implement MTPs, as well as some meso-level differences between the teachers from Norway and Russia regarding their reported implementation of MTPs. The first section contains the findings concerning the factors the participants reported as affecting their implementation of MTPs, while the second section covers the language-and country-related differences in the participants' implementation of MTPs that were discovered during data analysis.

Factors affecting the participants' implementation of MTPs
An analysis of the interview data led to the discovery of three overarching themes (see Figure 1) that were repeated across the interviews, regardless of how many (or which) languages the participants reported teaching, their age, gender, or any other variable. These themes were referenced consistently during the interviews and the participants came back to them often when reflecting on their ability and willingness to implement MTPs. Language knowledge encompasses the nature and extent of the participants' knowledge of the languages they taught, specifically, their metalinguistic knowledge and cross-linguistic awareness, and how they felt this knowledge affected their implementation of MTPs. Language learner represents how the participants positioned themselves as multilingual language learners, as well as their activities in support of this positioning, whereas support covers the help they reported receiving from various sources to implement MTPs. These three themes are described in detail in the following subsections.

Figure 1
The main themes and sub-elements linked to the participants' implementation of MTPs

Language knowledge
Seventeen participants (excluding Frida, Irina, Johanna, and Pia) referred to their language knowledge concerning the languages they taught when asked whether they drew on languages other than the target language during lessons. Except for Filippa and Vladimir, who reported possessing a high level of linguistic competence overall in each of the languages they taught, the other participants indicated shortcomings in their ability to verbalize language form, function, and meaning. Indeed, the data revealed that the participants' metalinguistic knowledge, more than any other aspect of their linguistic competence, played a role in determining the extent to which they implemented MTPs, which in many instances consisted of the participants drawing cross-linguistic comparisons (e.g., cognates) between languages and translating from one language to another.
The shortcomings reported by the participants with respect to the languages they taught were present regardless of whether they taught one or two FLs. For example, Siri, referring to her knowledge of the languages she taught, described herself as a teacher of English and an advanced learner of Spanish. She said that English was a much stronger aspect of her teacher identity and she regularly spoke English with her children. In contrast, she used Spanish primarily at school. She said she continued to learn Spanish in her free time to make sure that she could offer accurate feedback to students when called upon, although she added that her knowledge of Spanish sometimes failed her. However, despite her lower overall reported proficiency in Spanish when compared to English, Siri noted that her ability to verbalize language structure in Spanish was much stronger than it was in English. She reported regularly implementing MTPs in her Spanish lessons by pointing out cognates in English and Spanish (she said that both languages were strongly influenced by Latin), whereas she could not do the same in her English lessons: Maybe my biggest weakness as an English teacher is that I don't know enough of the technical names of the different grammar features because I don't use them that much while in Spanish, I use them a whole lot and I know more of the names and I know how to describe things better because I have had to learn them myself in the recent past and I use them every week in the classroom. (Siri) Filippa and Vladimir, in contrast to Siri and the other participants, expressed high confidence in their overall linguistic competence vis-à-vis the languages they taught, including in their metalinguistic knowledge of these. Their high levels of competence, especially their ability to verbalize language form, function, and meaning, led to them implementing MTPs across languages. For example, Vladimir reported that he had many of the same students in both his English and German lessons and that this allowed him to improve their knowledge of both languages simultaneously. He reported using a combination of English, Russian, and German in his lessons. Providing an example of how he drew on multiple languages, he explained that he taught his students how to guess the past tense of irregular verbs in both German and English by organizing activities that showed his students how both languages shared the same Germanic roots. He pointed out to them how English verbs like think, seek, and bring all adopted the ending -ought when conjugated in the past and how this was similar to their German counterparts denke (dachte), bringe (brachte), and suche (suchte), with the ght in English replaced by cht in German. The ght in English words like night and daughter similarly became cht (nacht and tochter) in German. He noted that the system was not perfect, working for some students while not for others, even though he considered it a useful strategy overall for predicting words and verb forms in both German and English. He also said that he drew his students' attention to the similarities between phrasal verbs in English and the German partikelverb (e.g., aufgeben 'give up,' aufheben 'pick up;' auf = up). He stated that his advanced knowledge of multiple languages had given him insights into how the languages were structured and used so that his English and German native speaker friends would turn to him with questions about their respective languages even though he was a non-native speaker: My German friends, when they want to know why a certain thing is the way it is in German, they ask me about it and, when I explain it, they are like . . . wow . . . we didn't even think about that. And it is almost the same with English. (Vladimir) As for Filippa, despite the significant differences between Chinese, English, and Russian, she felt that her knowledge of multiple languages allowed her to use a large range of teaching strategies to help her students learn more effectively. For instance, she reported translating from Chinese to Russian when teaching, as well as using her students' knowledge of word stress in Russian to help them master the various tones in Chinese. She said that most Russian words had specific stress patterns that she found useful when developing her students' Chinese pronunciation. She also reported drawing her students' attention to differences between Chinese and Russian regarding verb placement and the use of prepositions in sentences.

The teacher as language learner
Thirteen participants (excluding Anya, Elsa, Frida, Irina, Johanna, Pia, Rebecca, and Roman) explicitly positioned themselves as language learners when discussing their teaching practices. They stated that they were either learning new languages or working on improving their knowledge of the languages they taught, or a combination of the two, and acknowledged that these experiences had strengthened their implementation of MTPs and boosted their awareness of the effects of multilingualism on the language learning process.
The participants provided various reasons for learning new languages, with these reasons covering their love of languages, European identity, interest in a particular region of the world, and desire to understand and engage with their students more deeply. For instance, Kirill, a teacher of French from Moscow, revealed that he was learning German and, to a lesser extent, Turkish. He was learning Turkish because he was interested in the Middle East (he also stated that he had traveled to Tunisia to learn Arabic over two months in the summer a few years ago) while he felt that German complemented his knowledge of French and formed part of his European identity. He said that learning new languages had helped him hone his language learning strategies, which he had passed on to his students. These strategies included translanguaging, something that he saw as beneficial when learning new languages. He noted that he often consciously switched between languages outside of lessons when interacting with his friends, many of whom, he said, knew two or more languages. Learning German and Turkish had boosted his cross-linguistic awareness vis-à-vis French, and he added that he wanted to explore the connections between these languages more deeply. Siri, who drew on English to teach Spanish, said that she was learning Italian for personal reasons and that this, likewise, had had a positive impact on her teaching: "Like right now, as a grown-up, I learned Spanish and I am learning Italian and I feel like it's giving me a lot as a language teacher to be a learner myself" (Siri).
As for learning languages to understand their students better, the participants cited their interest in studying how language learning was enhanced through certain language combinations, as well as their wish to serve as a good multilingual model for students navigating multilingual, multidenominational, and multicultural environments. For instance, Astrid discussed her efforts to learn Arabic and Chinese as a way of understanding how her multilingual students used their knowledge of the other languages they knew to learn new languages: I have always liked languages and I've tried to learn Chinese several times because I have Chinese students. I also try to learn Arabic, and here it was more phonetics because I notice that my students who speak Arabic, they speak Spanish quite fast. They learn the phonetics pretty good. And so, I tried to learn Arabic to see how this happened. They also have a little bit of Greek [in Arabic]. So, they have something in the mix that helps to jump from one language to another. It is quite interesting. (Astrid) As was the case with the other participants, the learning of additional languages complemented Astrid's implementation of MTPs, which consisted of trying to understand how her students drew on their knowledge of Norwegian when making sense of Spanish and showing them how to translate between the two languages more effectively. In contrast to her Spanish lessons, she rarely used MTPs when teaching English (see Section 4.2.). Meanwhile, Svetlana, who compared English and Russian syntax during lessons and implemented translanguaging activities with her students, including translation, said that she was learning Arabic and Swahili to engage with her students more effectively and build rapport. She felt that it was important for teachers and students to understand one another given the increasingly multilingual and multicultural nature of language instruction, as well as for teachers to set an example for multilingual interactions: I want to learn Arabic as much as possible. I have students from Syria and when I open their notebooks, I see some Arabic writing and I like to communicate with them because we have some stereotypes and, to my mind, it helps us be more tolerant of other people. Now when we enter a classroom and see children of different nationalities, religions, we should give them, how to say, an example of how to behave and how to communicate to each other. (Svetlana)

Support
Seven participants (Astrid, Julia, Marte, Vladimir, Filippa, Svetlana, and Yevgeniya), all of whom had emphasized their language learning experiences during their interviews (see Section 4.1.2.), reported that they had received support regarding the implementation of MTPs when asked about their teaching practices. Such support, which was not systematic, occasionally came in the form of speaking with a colleague or mentor who made the participants more aware of the benefits of implementing MTPs, although most participants referenced their reading of research as having helped them become more aware of multilingualism as a pedagogical resource. For example, Vladimir, who taught English and German, reported reading up extensively on teaching cross-linguistically and credited the faculty at his university for giving him a sound foundation in both English and German pedagogy. He said he discussed cross-linguistic teaching strategies with his colleagues and felt that these exchanges had allowed them to share ideas and experiment with new teaching approaches. Similarly, Julia said that she had begun to implement MTPs more regularly during lessons thanks to a certain colleague who had helped her reflect more deeply on using her and her students' multilingualism as a resource when teaching. She also reported reading research on multilingualism in education. Asked if she thought her school environment encouraged her to implement MTPs, she noted that she was different from other teachers, whom she felt were sometimes not well-informed about the potential benefits of drawing on multilingualism as a pedagogical resource: It makes me a little frustrated to see all the levels of consciousness between the language teachers. I see a lot of problems and reflect on them. I think one thing might be that we don't talk. My advice to people is also to read, teachers need to work with some level of didactics in languages. (Julia) On occasion, she had organized joint lessons with a few other language teachers at her school, with the goal being to expose students to multiple FLs simultaneously. She felt that such a format was interesting and useful for boosting their language awareness. Like Julia, several other participants also cited a lack of awareness among their teacher colleagues regarding the benefits of implementing MTPs, as well as an absence of support from their school administration. Svetlana revealed that she did not receive any support from the school administration, nor was there a systematic procedure at school through which teachers could discuss experimenting with and refining their teaching approaches and practices. She felt that she was motivated to implement MTPs due to having read a lot, as well as due to the diversity in her classroom and her professional identity as an FL teacher: "I can't say I have always been interested. I read a lot and partly because of my profession and partly because I want to, I am really interested in differences. It is just, teachers should teach how to communicate." Marte, referring to the situation in her school, said that she did not generally discuss her implementation of MTPs with her colleagues because she felt that they did not know the languages they taught well enough, which affected their willingness to collaborate and reach out.

Country-and language-specific differences in the implementation of MTPs
Before presenting the findings concerning the country-and language-specific differences that were identified during data analysis, it is important to note that all 21 participants felt that being multilingual was an asset, even if three participants (Frida, Irina, and Pia) linked multilingualism to having a more positive multicultural outlook and did not see it as being essentially desirable when learning new languages. As for the language curriculum, this was described by several participants, especially Julia, as a guiding document that did not play a decisive role in how FL teachers taught day to day. Moreover, the participants, regardless of whether they taught in Norway or Russia, reported that their schools neither hindered nor encouraged them in their efforts to implement MTPs.
These similarities between the participants notwithstanding, there was one area where the participants from Norway and Russia differed in their implementation of MTPs: the teaching of English. Here, the differences between the Norwegian and Russian participants appeared to stem from the way English was taught in Norwegian and Russian schools. Twelve participants in total taught English (six from Norway and six from Russia). All the six Norwegian participants reported facing difficulties with the implementation of MTPs when teaching English, stating that they struggled to explain how the language was put together. For instance, when asked how confident she felt in her knowledge of English, Rebecca, one of the participants from Norway, responded that she found it lacking at times, especially when teaching immigrant students.
In middle school and especially in high school, you teach English, sort of politics, history, literature, and you don't teach the language anymore. And so, when people come in from south of Europe or the Middle East, then you have to teach them the language and you're not prepared to teach anyone the actual language. I don't feel as good at explaining why things are the way they are in English because I don't know the grammar well enough. (Rebecca) Even those participants that taught another FL alongside English reported possessing weak metalinguistic knowledge regarding the latter. For example, Siri could not implement MTPs in her English lessons as much as she reported doing in her Spanish lessons because she did not have access to a metalanguage in English that could help her accomplish this (see Section 4.1.1.). Astrid, another participant from Norway also reported drawing on Norwegian during her Spanish lessons to raise her students' cross-linguistic awareness, whereas she taught English mostly without implementing MTPs. In contrast, the Russian participants who taught English reported using translanguaging and drawing on cross-linguistic comparisons regularly during lessons. For example, Svetlana revealed that she used Russian with her students in English lessons not only to translate words but also to compare grammar concepts between the two languages. She explained that engaging in such explicit discussion of the differences between languages was normal for her and her students and that students in Russia studied Russian in schools with a similarly strong focus on one's explicit knowledge of the language: I think it is positive to use other languages because we may show some similarities and some differences. In my lessons, we often compare Russian and English. It is normal to compare different languages to see how they work and how people communicate. (Svetlana) Only Irina from the Russian group reported not using any MTPs during her lessons. She, along with Frida, was ambiguous about the benefits of using other languages in the classroom and supported an immersive approach where she only used the target language. According to Frida, using exclusively the target language allowed her students to use it more often in the classroom. She emphasized that their goal was to speak French in the classroom, not other languages.

Discussion
This study explored: (a) the factors that the participants consistently reported as influencing their implementation of MTPs and whether there were (b) language-or (c) country-specific differences in this respect. In terms of the first research question ("What factors do the participants consistently report as influencing their implementation of MTPs?"), the findings revealed three factors that the participants reported as influencing their implementation of MTPs: their language knowledge, the language learner aspect of being an FL teacher, and the support they had obtained to implement MTPs. Regarding their language knowledge, those participants who possessed advanced metalinguistic knowledge in the languages they taught reported facing few difficulties when implementing MTPs, regardless of their overall linguistic competence (for a discussion of linguistic competence, see Lehmann, 2007). In contrast, those participants, who reported shortcomings in their metalinguistic knowledge in a particular language, found it difficult to implement MTPs in that language. Several researchers have emphasized the importance of possessing advanced metalinguistic knowledge for teachers to successfully implement MTPs (Hofer & Jessner, 2019;Otwinowska, 2014), and so the findings regarding the participants' metalinguistic knowledge affecting their implementation of MTPs are not unexpected. Nevertheless, this study represents one of the few instances where FL teachers' metalinguistic knowledge has been explored as part of their ability to implement MTPs, and the findings help us pinpoint specific areas of FL teachers' linguistic competence that might require greater attention in teacher education programs if they are to successfully implement MTPs. In addition, the findings add greater depth to those reported in the study by Calafato (2021b), where statistically significant, positive correlations between FL teachers' implementation of MTPs and their self-assessed ability to teach diverse language aspects and skills were found, even though metalinguistic knowledge was not explicitly explored.
What is new in this study is that the findings indicated that, in the case of the participants who taught two FLs, metalinguistic knowledge in one FL did not always contribute to the ability to implement MTPs in the other FL. For instance, Siri implemented MTPs when teaching Spanish but could not transfer such practices over to her English lessons. As already mentioned, since most studies on FL teachers' implementation of MTPs have not investigated those teaching more than one FL, it is difficult to determine why some of the participants could not benefit from metalinguistic transfer between their languages. One possible explanation for the lack of transfer could be that they had a level of multicompetence where their languages were interconnected in their minds in such a way as to allow unidirectional metalinguistic transfer (i.e., only from one language to another) but not bidirectional transfer. Unidirectional transfer has been reported in studies on multilingual language learners concerning, for example, their morphological awareness (Ke & Xiao, 2015). Another possibility is that certain aspects of their languages had not been integrated, creating difficulties for the participants when it came to implementing MTPs. Cook (2016), in discussing variation in an individual's multicompetence, notes that all multilingual language users are on a continuum between a complete separation of languages and a complete integration of languages in the mind and that this separation or integration may vary from language aspect to language aspect. This study's findings suggest that, at a minimum, the metalinguistic aspect of their languages had not fully integrated for some participants, preventing them from carrying their implementation of MTPs in one FL over to the other FL. These dynamics highlight the complexity and diversity of interactions that occur between languages in the minds of FL teachers tasked with teaching two FLs and underscore the need for more studies that explore the cognitive aspects of their multilingualism.
At the same time, in this study, two teachers, Vladimir and Filippa, stood out in terms of their reported implementation of MTPs in the FLs they taught, and the findings suggested that they had advanced metalinguistic knowledge in these and a high degree of integration vis-à-vis their languages that allowed them to draw on their multilingual affordances (Aronin, 2014) in creative ways. For instance, Vladimir reported how he developed his students' cross-linguistic awareness and metalinguistic knowledge by teaching them strategies to identify patterns when conjugating the past tense of irregular verbs in English and German. He also reported carrying content taught in English lessons over to his German lessons and vice versa, which is similar to the strategy employed by Mr. Guo, the teacher of English and Chinese investigated by Jiang et al. (2014), who similarly integrated the content he taught in his Chinese and English lessons. Filippa, in contrast to Vladimir, reported teaching Chinese and English, languages that are much more distant from each other than are English and German. Yet, despite this distance, she reported implementing MTPs in both languages. This was evident in how she reported drawing on Russian, a language that she did not teach, to help her students learn the various tones in Chinese and sought to raise their awareness of the differences between Chinese and Russian regarding verb placement and the use of prepositions. These practices indicate that her multilingualism provided her with several possibilities for action so that she drew on additional (albeit distant) languages in creative ways to focus on language syntax and phonology cross-linguistically.
In addition to language knowledge, the findings also revealed that a majority of the participants positioned themselves as, and benefitted from being, language learners when discussing their implementation of MTPs. They reported learning new languages, further developing their language skills in the languages they taught, and even learning the languages their students spoke. As McDonough (2002) has noted, there have been few studies that have focused on teachers as learners, yet the learner aspect might be critical in obtaining a more holistic understanding of FL teachers, their relationship with the languages they teach, and their overall identity as multilingual language users. This is because FL teachers may not subscribe to the view of teaching as coming "from the 'top down,' as knowledgeable or skilled individuals seek to impart knowledge or skills" to their learners, the "ignorant or unskilled individuals" who "seek to become more knowledgeable or skilled" (Tomasello, 1999, p. 523). Rather, they may see benefits in being, and identifying as, language learners themselves, especially in terms of their ability and willingness to implement MTPs and teach more effectively. Several participants in this study considered being a language learner an integral part of their teaching and professional development, and as boosting their engagement with their students. The findings, like the ones concerning their language knowledge discussed above, call for a more nuanced look at FL teachers and underline the importance of exploring how they position themselves in relation to the languages they teach. The findings also underscore the need to look at teacher identity through a multi-faceted approach that takes into account how various micro-, meso-, and macro-level factors may influence their thoughts and actions (Douglas Fir Group, 2016;Gee, 2001). Support, meanwhile, mostly consisted of the participants reading research and, more rarely, collaborating with a colleague or mentor. The participants indicated that support was not systematically provided by their schools and, instead, depended on their individual efforts, which might explain why only a small number of them reported receiving it. The lack of systematic support from schools reported by the participants in this study echoes the findings from Calafato (2021b), where a majority of the FL teacher participants surveyed stated that their schools had not organized anything to promote multilingualism. The findings also reflect the studies reviewed in Calafato (2019), where schools either did not support the implementation of MTPs or actively opposed it in favor of a monolingual teaching approach. Moreover, some participants noted that many of their colleagues did not have a good understanding of how they could draw on their and their students' multilingualism as a pedagogical resource, and that collaboration did not occur because teachers did not talk to each other. Another reason reported by the participants for the lack of collaboration was that some teachers felt anxious about their knowledge of the FLs they taught and chose not to engage with their colleagues as a result. For teacher educators and educational institutions, these findings provide insights into FL teacher interactions and their potential to collaborate with each other. They also highlight the importance of developing teachers' language ability alongside their pedagogical knowledge since the former was found to not only influence their implementation of MTPs but also affect their willingness to collaborate with each other, this being an important component of multilingual pedagogy (Haukås, 2016).
Finally, with regard to the second ("Are there differences in the participants' reported implementation of MTPs based on the languages they teach?") and third research questions ("Are there country-specific differences in the way the participants report implementing MTPs?"), the findings did not indicate any notable differences between the teachers based on the languages they reported teaching, although differences were found between the participants from Norway and Russia concerning the teaching of English. The lack of language-specific differences in the participants' implementation of MTPs should be considered in light of the three factors that they consistently reported as affecting their ability and willingness to implement MTPs (especially the language knowledge factor). In other words, it appears that language knowledge, and not language type, affected their implementation of MTPs. The findings, in some ways, differ from those reported in Calafato (2021b), where statistically significant differences in MTP implementation were found based on what languages the participants taught. At the same time, it is worth keeping in mind that not all FLs were equally represented in this study. For example, more participants taught English than Chinese. In addition, a small number of participants overall were interviewed in the study, and the findings were based on teachers' self-reports and not classroom observation.
As for the third research question, participants from Norway, without exception, reported possessing weak metalinguistic knowledge in English and found it difficult to implement MTPs in their English lessons. Studies indicate that students in Norway learn English mostly implicitly, without a strong focus on metalinguistic learning strategies (Haukås, 2016), something that was corroborated in this study. Consequently, the participants from Norway who taught English might not have sufficiently developed their metalinguistic knowledge in the language as language learners, and their enrollment in teacher education programs as adults may have failed to offset these years of implicit learning. The participants from Russia who taught English, in contrast, did not report issues with their metalinguistic knowledge in the language and stated that this was how languages were generally taught in schools in Russia, including Russian. The findings imply that, with respect to the participants from Russia, their metalinguistic skills were being developed in both an FL and the country's official language (i.e., Russian) already from a young age, whereas this was not the case with the participants from Norway, who would have waited until they started learning their second FL at school to begin developing their metalinguistic knowledge (see Haukås, 2016).

Conclusion and implications
This study is one of the very few to have investigated the factors that FL teachers report as influencing their implementation of MTPs across multiple FLs (i.e., Chinese, English, French, German, and Spanish), including in instances where they taught more than one FL. The study's findings revealed that the participants' metalinguistic knowledge, more than their overall linguistic competence, affected the extent to which they implemented MTPs. Moreover, such metalinguistic knowledge did not always transfer between languages, likely as a result of participants possessing a level of multicompetence where, at the very least, the metalinguistic aspect of their languages had not been integrated. A majority of the participants also positioned themselves as language learners, which they felt had a positive effect on their teaching practices, notably their implementation of MTPs. At the same time, a limited number of participants reported receiving support for implementing MTPs. This support was not systematic and mostly involved reading research reports. Collaboration with colleagues and access to mentors occurred more rarely.
The findings contribute to our understanding of the factors that FL teachers report as influencing their implementation of MTPs, with strong implications for teacher educators, teacher education programs, and future research on multilingualism as a pedagogical resource. First, they underline the need for educational institutions and teacher educators to ensure that teacher education programs place sufficient emphasis on assessing and developing FL teachers' language knowledge, in much the same way as they do for their pedagogical knowledge and general awareness of multilingualism in the language classroom. Indeed, a focus on pedagogical knowledge and awareness of multilingualism alone will likely prove insufficient in getting FL teachers to implement MTPs if they feel that their language knowledge is inadequate for the task. Specific measures that teacher educators can use to develop teachers' language knowledge include requiring them to take courses that cover the pedagogical applications of metalinguistic knowledge and cross-linguistic awareness, as well as the activities that they can devise to develop these in their students through the use of an appropriate metalanguage. This might require teachers to become familiar with linguistic terms in a way that connects these terms more explicitly to their teaching. Teacher educators can also encourage their students to reflect more deeply on their multicompetence and affordances, especially the extent to which various language aspects are integrated across the languages they know and/or will teach. Such reflection would help them pinpoint specific areas for improvement and could ultimately lead to more effective MTP implementation.
Second, the focus on developing FL teachers' language knowledge should form part of a larger push towards a reconceptualization of FL teachers, one that takes into account that they often view themselves as, and see benefits in being, language learners. What this means is that educational institutions should provide FL teachers with avenues through which they can nurture the language learner aspect of their identity. For instance, teacher education programs could require FL teachers to learn a new language to a certain level of competence or conduct research on a language that is spoken by a large number of students (but which the teachers know little about) as part of their studies. In many ways, the MTPs that FL teachers are encouraged to use with their students could be employed by teacher educators to train pre-service teachers. Educational institutions could also organize regular workshops and seminars where FL teachers would be able to reflect on their language learning experiences, engage in pair and group tasks to pool together the lessons they learned from these experiences, and find ways of incorporating them more effectively into their teaching.
Third, teacher education programs and educational institutions should stress the importance to FL teachers of building a support network of colleagues with whom they can discuss their approaches to teaching FLs and the possibilities of implementing MTPs. In this respect, schools and universities could organize tandem teaching programs, where they would pair up novice FL teachers with more senior teacher colleagues who have experience implementing MTPs. The tandem could last a semester, an academic year, or several years, and involve the pair discussing and planning lessons together, mentorship sessions, and lesson observations, among other activities. The tandem would end with the pair writing a reflective report on their experiences, the improvements they made to their teaching as a result of the program, and the areas that they felt still required improvement.
Fourth, the findings underscore the need for additional studies on teachers of multiple FLs, specifically, observation studies that would explore differences in their teaching practices per FL and the reasons for these differences, including the planning that goes into preparing lessons per FL. In this study, the findings derived from the participants' reported practices and it is possible that what they reported doing differed notably from what they actually did in the classroom. Future studies should also cover diverse language combinations, that is, teachers of languages from the same group (e.g., English and German), as well as those teaching languages from distant groups (e.g., Chinese and Spanish), ideally using a large participant sample.
Fifth, as this study has shown, it is important to differentiate between various types of multilingualism, between FL teachers who might be multilingual but teach one FL and those who are multilingual yet teach two or more FLs. In addition to such differentiation, the findings highlight the importance of exploring a range of micro-, meso-, and macro-level factors (not just classroom teaching experiences) when seeking to understand how FL teachers relate to their multilingualism and the extent to which they are willing and able to implement MTPs. These factors can include the attitudes of the school administration, the perceived importance of the curriculum, and interactions with colleagues and students. Finally, researchers could look into how the teaching of FLs alongside non-language subjects affects teachers' implementation of MTPs, including if they carry the MTPs they use in their language lessons over to the non-language subjects they teach and whether the teaching of language and non-language subjects side by side leads to changes in how they conceive of language as an epistemic tool.